While the prevalence of childhood obesity seems to have reached a plateau recently, substantial disparities in child obesity by race/ethnicity and family socio-economic status (hereafter, social disparity in child obesity) continue to widen; the rates decreased in non-Hispanic white or high-socioeconomic status (SES) children but increased among minority or low-SES children. Given the adverse health outcomes that result from obesity, it is critical to examine which factors affect childhood obesity and its social disparity. Changes in neighborhood SES over the last 2 decades correspond to the recent increases in social disparity in child obesity, thus neighborhood SES is potentially one important contributor to the greater social disparity in child obesity. Past studies have established a scientific premise for research in this area; however, the majority of evidence has (1) been based on cross-sectional studies, (2) not accounted for residential mobility of individuals, (3) utilized less valid self-reported heights and weights, and/or (4) not examined mediating mechanisms by which neighborhood SES affects child obesity. Further, little research has examined whether neighborhood SES has differential effects on child obesity by important subgroups (e.g., based on race/ethnicity and family SES). This proposed study is a secondary data analysis of a large, longitudinal study, namely the Fragile Families & Child Wellbeing Study (FFCWS), to examine the longitudinal association between neighborhood SES changes and childhood obesity and its social disparity. The FFCWS is a $40 million study of 4,898 children born in 1998- 2000 and their families. This proposed study will address the gaps described above by using data from FFCWS. The specific aims are to examine from young childhood to adolescence: 1) the contribution of neighborhood SES changes on child obesity; 2) the extent by which neighborhood SES changes explain social disparity in obesity; 3) the mediating effects of individual obesity-related behaviors and neighborhood social environments on the association between neighborhood SES changes and child obesity; and 4) the interacting effects of individual sociodemographic factors on the association between neighborhood SES changes and child obesity. Using the FFCWS is advantageous because data from the FFCWS (1) include objectively- measured body mass index of children when they were three, five, nine, and fifteen; (2) have information about multiple levels of exposures ranging from the individual and family levels to the neighborhood and social levels, which can be utilized as potential moderators, mediators, or covariates; and (3) were collected in 2000s during the drastically changing time period in neighborhood SES, which would allow for better causal inference. This study builds on important preliminary studies and the extensive relevant experience of two principal investigators. This study is very cost-effective and will elucidate the role of neighborhood socioeconomic status on child obesity including mediating and moderating relationships, allowing for recommendations that will inform community-level interventions to reduce child obesity.